Public Health Knowledge Management – One size does not fit all (paper by Sarah Harlan et al.)

(picture copyright by DFID on flickr)

In a recently published paper about knowledge management in family planning in Ethiopa, Sarah Harlan (of Johns Hopkins) et al. dig deep into the knowledge networks that connect the national and local government agencies with women and men on the ground. They combined Net-Map with focus group discussions and other interviews. Among their findings is the insight that actors on different levels have different knowledge needs and these are best supported by different network structures. I am especially interested in how high centralization impacts on a networks ability to deliver.

No network structure is perfect and every structure has pros and cons. If you have a highly centralized knowledge network, basically Ministry of Health in the middle, distributing information to everyone, this can be great in some respect, because you are in control of the message. In a context such as reproductive health, where traditional beliefs and modern medicine might be contradictory, having one central distributor of information can help to make sure the message that local women receive is consistent. Also, centralized systems are easy to understand for people within and outside the system, users know who to turn to, donors know whom to support. However, centralization also has a number of risks: They put a high burden on the central node – if the Ministry of Health is overloaded or does not perform well, nothing will happen. Also, they are weak at producing and sharing locally adapted solutions. So, while they work well in situations where there is one clear correct answer, they are weak, slow and not very creative in situations where many different solutions will work. For example, if the question is: How do we get village women to give birth in a hospital instead of at home, there can be many different solutions that work and that could inspire others – if you have a system that is designed to share information in a decentralized manner.

Do you have a Net-Map paper, report, blog post that you want to share? Send me a note and I am happy to include you here.

The Network of Europe’s E. Coli (EHEC) Crisis

Does this look like a killer? (picture by yogendra174 on flickr)

Network analysis is great for getting some clarity and a sense of direction in complex, messy, multi-actor situations, and I guess that is a pretty good description of the situation in Europe’s E. Coli outbreak (just how messy? These CNN and Newsweek articles give some insight). It’s a multi-actor situation, the actors involved include different government departments (e.g. for health, agriculture, trade) of different countries (Germany, where the outbreak happened, other countries that may or may not have produced the vegetable responsible) and at the EU level, everyone in the supply chain, from farmers to traders and food businesses to consumers, research organizations, media outlets etc.

But what makes this really messy, is the fact that these actors are linked by very different kinds of connections. There are the material flows of produce (infected or not) from farm to table. There is the movement of the infection through the system, with the bacteria sometimes connected to the produce, but now, further into the outbreak, more likely a person-to-person connection. As if this wasn’t complex and obtuse enough, the way that countries and the EU as a whole deal with this issue is structured by the administrative networks and hierarchies. And because multiple departments or ministries in multiple countries are involved, this is not just one hierarchical pyramid, where the person in the top position can decide what to do, but a number of (internally hierarchical) organizations, which are linked to each other by non-hierarchical links and somehow have to figure out how to collaborate quickly and effectively once new information becomes apparent.

And this leads us to the fourth kind of link we need to look at, to understand how this outbreak works and that is the information link: Where is new knowledge about sources, spread, effects and treatments produced and how does this knowledge reach those who matter, who tells decision-makers what they need to know to make good decisions, who informs (and doesn’t mis-inform) the media, where does the general public get their information?

Now imagine you sat down with a group of researchers, government and farmer representatives to map out the actors, how the produce flows between them, how the infection has spread so far, how they are connected in terms of hierarchy and coordination and where the information flows through the system. The resulting map might be so big and messy that it first blows your mind. But (even without mapping it) that is the reality that the decision makers on the ground have to deal with; under time pressure and with high stakes.

And as we have seen when mapping out the response to bird flu outbreaks, just putting this map together can help those involve discover crucial oversights and structural problems in the system. In Ghana, we were able to point out how the structure of the compensation schemes created corruption hot spots at the border and where structural holes led to a risk of communication breakdowns.

However, when reading about the outbreak, certain network links are in the forefront of everyone’s attention: The big question everyone is asking is: How does the infection travel through the system (did it ride a cucumber or a bean sprout)? And while this is one crucial question, the questions that will have a greater long term impact are: How ready was the system to deal with the outbreak? How did information flow and how are interventions coordinated? Jack Ewing hints at that in a New York Times article, pointing out that the federal decentralized system of disease reporting led to a crucial delay in understanding the scope of the threat. But to increase readiness for the outbreak of any contagious disease (be it the swine flu or bird flu or a new strain of E. coli) it is crucial to improve the information and intervention systems along with discovering the source of a specific outbreak.